Hyperglycemia and Outcome in the Pediatric Intensive Care Unit

Objective

To identify the frequency of hyperglycemia in children who are nondiabetic and critically ill and assess the independent effect of hyperglycemia on outcome.

Study design

Consecutive admissions to the pediatric intensive care unit (PICU) were reviewed. The Pediatric Risk of Mortality III score (PRISM) measured patient acuity. Because maximum glucose level in the first day of PICU admission (GLFD) >200mg/dL contributes to PRISM, 200 mg/dL was used to differentiate high glucose (HG) from normal glucose.

Results

Of 1550 patients, 221 (14.3%) had HG. GLFD correlated with PRISM (r = 0.39, P < .001). Without controlling for PRISM, the HG group had more mechanical ventilation days (MVD; P < .001), longer PICU length of stay (PLOS; P < .001) and lower percent survival (P < .001) than the normal glucose group. Controlling for PRISM in survivors, GLFD was not associated with PLOS (P = .75) or with MVD (P = .06). GLFD was not significantly associated with survival (P = .76). In nonsurvivors, GLFD was not associated with PLOS (P = .19) or MVD (P = .31).

Conclusion

When controlling for disease severity, hyperglycemia within 24 hours of PICU admission was not independently associated with increased mechanical ventilation time, length of stay, or mortality. Prospective evaluation of glycemic control in critically ill children is needed to elucidate its effects on outcome.

Abbreviations: ANCOVA, Analysis of covariance, GLFD, Maximum glucose level in the first day of PICU admission, HG, High glucose, ICU, Intensive care unit, LOS, Length of stay, MVD, Mechanical ventilation days, NG, Normal glucose, PICU, Pediatric intensive care unit, PLOS, PICU length of stay, PRISM, Pediatric Risk of Mortality III